LNG-fuelled deep sea shipping

LNG-fuelled deep sea shipping The outlook for LNG bunker and LNG-fuelled newbuild demand up to 2025 August 2012

Cover image: Viking Line s LNG-fuelled ferry, Viking Grace, under construction at STX Turku in Finland. Image courtesy of STX Finland. Lloyd s Register Group Limited, its affiliates and subsidiaries and their respective officers, employees or agents are, individually and collectively, referred to in this clause as Lloyd s Register. Lloyd s Register assumes no responsibility and shall not be liable to any person for any loss, damage or expense caused by reliance on the information or advice in this document or howsoever provided, unless that person has signed a contract with the relevant Lloyd s Register entity for the provision of this information or advice and in that case any responsibility or liability is exclusively on the terms and conditions set out in that contract. Copyright Lloyd s Register Group Limited. 71 Fenchurch Street, London EC3M 4BS, 2012. Except as permitted under current legislation no part of this work may be photocopied, stored in any medium by electronic means or otherwise, published, performed in public, adapted, broadcast, transmitted, recorded or reproduced in any form, without the prior written permission of the copyright owner. Enquiries should be directed to the above address. Where Lloyd s Register has granted written permission for any part of this publication to be quoted such quotation must include appropriate acknowledgement to Lloyd s Register.

Contents 1 Foreword 3 2 A summary of the key study findings 4 3 Marine bunker fuels and sulphur content regulation 6 Marine bunker fuels 6 Stricter sulphur content limits in marine bunker fuels 6 Compliance with stricter limits on fuel oil sulphur content 7 4 Study of LNG as a fuel for deep sea shipping 9 Bunkering infrastructure: a barrier to LNG adoption 9 The objectives of the study 9 The study steps and limitations 9 5 LNG bunker demand assessment shipowners survey 11 Survey findings 11 Main bunkering ports for deep sea ships 11 Shipowners intentions for mitigation of SOx emissions 11 6 LNG bunker supply assessment port survey 13 Survey findings 13 Awareness of LNG as fuel as a compliance option 13 Plans for LNG bunkering infrastructure 13 7 Forecast of LNG-fuelled newbuild and bunker demand 15 LNG bunker demand model 15 The three demand scenarios: base case, high case and low case 16 LNG bunker and newbuild forecast: base case scenario 16 LNG bunker and newbuild forecast: high case scenario 19 LNG bunker and newbuild forecast: low case scenario 22 8 Conclusions 26 9 References 27 10 Abbreviations and terms 27 11 Figures and tables 28 Appendix 1 Port survey analysis data and results 29 Top 10 global bunkering locations 29 Selection of bunkering ports for survey 29 Additional notes on specific port selection 31 Port profiles 31 The survey 32 Part 1 Awareness 32 Part 2 Planning for LNG bunkering 34 Appendix 2 LNG bunkering newbuild demand model 40 Shipping trade route inputs 40 Regulatory drivers 42 LNG supply route assumptions 43 Vessel deployment and operational factors routes/vessels assumptions 44 Comparable fuel prices of LNG versus HFO with scrubber and MGO 44 Engine and abatement technology cost assumptions 45 World fleet development by shiptypes 1991-2025 45 Global newbuild forecasts for deep sea trades 2012-2025 47 Global newbuild forecasts for deep sea trades 2012-2025 48 Global LNG bunker consumption demand 2012-2025 51 1

2

1. Foreword There is a heightened focus on reducing emissions from shipping to minimise the impact of air pollution on the environment, driven by regulations and stakeholder expectations. Low-sulphur alternative fuels are one option being considered to achieve this reduction, and this has in turn led to serious consideration of liquefied natural gas (LNG) as a marine fuel. In April 2011, we commissioned a study to understand how a global LNG bunkering infrastructure might develop and to assess the likelihood of LNG being widely adopted as a fuel for deep sea shipping. From this study, we have: 1. identified strategic ports and locations worldwide for possible LNG bunkering infrastructure facilities, and gathered the opinions of bunkering ports on their likely provision of LNG bunkering facilities in future 2. assessed the likely scale of demand for LNG-fuelled new construction and LNG as a fuel for deep sea shipping up to 2025, using a proprietary interactive demand model. The results of this demand model, which is based on three possible future scenarios for LNG pricing and implementation of global sulphur limits, show that competitive pricing of LNG could see the fuel adopted widely for deep sea shipping. The model has also helped us understand there is a fine balance of influencing factors, such as fuel price differentials and the timing of global sulphur limit enforcement, that will influence the industry s decisions about future fuels. Making the model available is enabling us to work with many industry stakeholders, including ports, shipowners, shipbuilders and gas suppliers, and we will continue using it to provide consultative services for LNG fuelled newbuilds and LNG bunkering. Our thanks go to Maritime Strategies International Ltd (MSI), a London-based shipping economic consultancy firm, who helped us carry out the study. Hector Sewell Head of Global Marine Business Development Lloyd s Register 3

2. A summary of the key study findings Marine bunker fuels and regulation of sulphur content Heavy fuel oils (HFO) with high-sulphur content accounted for 76% of marine bunker fuel demand in 2010. To limit emissions of the harmful pollutant sulphur dioxide (SOx) from ships, strict limits on sulphur content in marine bunker fuel oils are being implemented in coastal areas known as Emission Control Areas (ECAs). A strict global sulphur content limit of 0.5% could also be implemented in 2020. As the schedule for the sulphur limits approaches, LNG as bunker fuel is being considered as one alternative to conventional marine bunker fuel oils because it produces emissions with a SOx content of virtually 0%. LNG bunker demand assessment shipowners survey From a survey of shipowners on deep sea trades: Low-sulphur fuel oil is seen as a short-term option for compliance with SOx emission regulations. Abatement technologies are seen as a medium term option. LNG-fuelled engines are a viable option in the long term, particularly for ships on liner trades. LNG bunker supply assessment port survey From a survey of bunkering ports: LNG bunkering is expected for short sea shipping in ECAs. LNG bunkering may eventually cascade into deep sea trade facilitated by regulations. LNG bunker demand is highly dependent on LNG pricing and its comparable price difference with competing fuels, for example HFO and marine gas oil (MGO). Forecasts of LNG fuelled newbuild and bunker demand Using the LNG bunker demand model, three scenarios have been developed and examined based on assumptions for: wider implementation of ECAs the date of the strict global sulphur limit implementation the propensity of shipowners to adopt LNG as a fuel for newbuilds bunker fuel oil and LNG bunker price forecasts. The three forecast scenarios for LNG-fuelled newbuild and LNG bunker demand are: Base case scenario current ECAs and a 0.5% global sulphur limit in bunker fuel implemented from 2020: 653 LNG-fuelled newbuilds forecasted for the period up to 2025 (4.2% of global deliveries from 2012 to 2025) LNG bunker demand is expected to reach 24 million tonnes (MnT) by 2025 for deep sea trades (1.5% of global LNG production and 3.2% of global HFO bunker consumption). 4

High case scenario a 25% decrease on the forecast LNG bunker prices used in the base case model and a 75% increase in propensity for newbuilds to convert to LNG-fuelled designs from 2020-2025: 1,963 LNG-fuelled newbuilds forecasted for the period up to 2025 (12.6% of global deliveries from 2012 to 2025) LNG bunker demand is expected to reach 66 MnT by 2025 for deep sea trades (4.2% of global LNG production and 8.0% of global HFO bunker consumption). Low case scenario a 25% increase in forecast LNG bunker prices used in the base case model and implementation of global sulphur limits shifting to 2023. Sensitivity testing indicates that shifting implementation to 2025 for the low case would generate a zero demand for LNG-fuelled newbuilds: 13 LNG-fuelled newbuilds forecasted for the period up to 2025 (0.1% of global deliveries from 2012 to 2025) LNG bunker demand is expected to reach 0.7 MnT by 2025 for deep sea trades (0.001% of global LNG production and 0.002% of global HFO bunker consumption). LNG-fuelled deliveries base LNG-fuelled deliveries high LNG-fuelled deliveries low LNG bunker consumption base (RH axis) LNG bunker consumption high (RH axis) LNG bunker consumption low (RH axis) 2,500 80 69.6 70 2,000 1,963 60 Number of ships 1,500 1,000 24.1 50 40 30 LNG bunker (MnT) 500 653 20 10 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0.7 13 0 Figure 1: Cumulative global LNG fuelled newbuilds and LNG bunker consumption (base, high and low cases) 5

3. Marine bunker fuels and sulphur content regulation The most widely used marine bunker fuel is heavy fuel oil with high sulphur content, but regulations are implementing strict limits on sulphur content in marine bunker fuel oils to limit emissions of sulphur dioxide from ships. Marine bunker fuels Currently there are three basic types of marine bunker fuel on the market: Residual fuel oil this is the heaviest oil fraction from the oil refining processes and is often called heavy fuel oil (HFO). It is the traditional marine bunker fuel and has a high sulphur content. Distillate fuel oils these are lighter oil fractions from the oil refining process. They typically have a low sulphur content. Intermediate fuel oils (IFO) these are a mixture of residual and distillate fuel oils. Within this report, MGO is used as a general term for all low-sulphur fuel oils because the price difference between the different types of low-sulphur fuel oils (MDO, MGO, etc.) is negligible when compared to the price difference between fuels like HFO and LNG. Global marine bunker demand was estimated to be 235 MnT in 2010 (IEA, 2010) of which HFO accounted for 76%. Demand for intermediate and distillate fuel oils was highly concentrated around current ECAs and was negligible elsewhere. Stricter sulphur content limits in marine bunker fuels The International Maritime Organization (IMO) has adopted measures to prevent air pollution from ships through MARPOL Annex VI. From 2012, MARPOL Annex VI includes a global cap of 3.5% on the sulphur content of marine bunker fuel to limit emissions of sulphur dioxide, a harmful pollutant. From 1 January, 2020, a global sulphur limit of 0.5% in marine bunker fuel oils is expected. However, this date could be deferred to 1 January, 2025, depending on the outcome of further investigation by the IMO into the global availability of low-sulphur fuel oil for marine use by 2018. 1.1.2012 4.50% 1.1.2020* 1.1.2025 3.50% Fuel oil sulphur limits 1.50% 1.00% 0.50% 0.10% Outside ECA-SO x 1.7.2010 1.1.2015 Inside ECA-SO x *Depending on the outcome of a review of fuel oil availability, to be completed 2018, the 2020 date could be deferred to 2025 Figure 2: MARPOL Annex VI regulations and enforcement of sulphur limits with respective timelines 6

Since 2005 a number of Emission Control Areas (ECAs) have been implemented (see Table 1). The sulphur content limit of marine bunker fuel in ECAs is far more stringent. The current limit is 1%, reducing to 0.1% by 2015. Current and confirmed Emission Control Areas (ECAs) Entry into force* Baltic Sea (SOx) 19 May, 2005 North Sea (SOx) 22 November, 2006 North America, US including Canadian coast up to 200 nautical miles (SOx, NOx and PM) US Caribbean Sea covering Puerto Rica and US Virgin Islands ECA (SOx, NOx and PM) 1 August, 2011 1 January, 2013 * Stricter limits for fuel oil sulphur content are applied one year after the date of entry into force Table 1: Current and confirmed emission control areas Compliance with stricter limits on fuel oil sulphur content Operating on low-sulphur distillate fuels is a relatively easy way to comply with fuel oil sulphur content limits. However, if the world fleet of commercial ships was to convert to using distillate fuel by 2020, the current production of distillate fuel oil would not meet marine bunker fuel demand. According to a recent report (Meech, 2011), Outlook for Marine Bunkers and Fuel Oil to 2030, the refinery industry would need to produce an additional 4 million barrels per day of distillates in order to meet demand for bunker fuel oils for shipping on implementation of the 2020 IMO global sulphur limits. Doubts about availability, and pricing, of distillate fuels have led to other options for compliance with strict sulphur limits being considered. The three main options are: 1. operating on low-sulphur fuel oil (LSFO), which, depending on sulphur content limitation, could be marine diesel oil (MDO) or MGO (i.e., distillates) 2. operating on HFO with an exhaust gas treatment system (EGTS) 3. operating on liquefied natural gas (LNG). A comparison of these three main options is shown in Table 2, along with corresponding reductions for all types of current and expected future regulated emissions, i.e., CO 2, SOx, NOx, and particulate matter. Use of LNG as bunker fuel for ships represents a real alternative to conventional marine bunker fuel oils when considering compliance with more stringent sulphur limits because of its virtually 0% SOx content in emissions (depending on engine type). 7

Compliance option LNG HFO MDO/MGO CO 2 removal 10-20% SOx removal 100% NOx removal Up to 80-90% Particulate matter 98-100% Abatement technologies No MDO: <2%; MGO: 0.01-1% Abatement technologies Regulation in place Developing Yes Yes Infrastructure Early stages Yes Yes Cultural factors Higher Established Established Cost of use Potential to stretch the technology Challenges /differences LNG storage tank size; LNG fuel price uncertain; possible loss of cargo space Further CO 2 reduction Bunker space/cryogenics /possible methane slip Abatement technologies required End of cycle Abatement technologies Varied blends of distillates 2020 Table 2: The three main options for compliance and corresponding emission reductions 8

4. Study of LNG as a fuel for deep sea shipping In April 2011, we commissioned a study to understand how a global LNG bunkering infrastructure might develop and to assess the likelihood of LNG being widely adopted as a fuel for deep sea shipping. Bunkering infrastructure: a barrier to LNG adoption A critical aspect of the development of LNG as a fuel is the lack of an established bunkering infrastructure and supply chain network for delivering LNG as a marine fuel. This is considered a significant barrier to the widespread adoption of LNG as a fuel, with gas providers and bunker suppliers unwilling to invest in the necessary infrastructure until there is sufficient demand to supply commercial shipping with LNG fuel. On the other hand, shipowners are unwilling to invest in LNG-fuelled ships if supplies of LNG bunker are difficult to obtain. The objectives of the study In view of the significance of this bunkering infrastructure to the development of LNG as a marine bunker fuel for deep sea shipping, the study had two main objectives: 1. To survey industry stakeholders (shipowners and ports) to understand how they would respond to the stricter sulphur content limits in the short, medium and long-term, including the likelihood of ports developing LNG bunkering infrastructure and shipowners adopting the option of LNG as a fuel. 2. To create a proprietary interactive model of future demand for LNG-fuelled deep sea ships and volumes of LNG they would consume on deep sea trade routes, and to use this to create forecasts of the likely take up of LNG as a fuel for deep sea shipping to 2025. The study steps and limitations The study comprised a number of steps and tasks, shown in Figure 3. The following aspects were not explored in detail since they would require a separate study: LNG-fuelled ship technology, including energy density, specific gravity and tank volumes per fuel option onshore LNG bunkering technology, including bunkering standards the implications of other emissions regulations for bunker fuel choice, such as CO 2, NOx and particulate matter regulations other possible fuel options, including LPG, bio-fuels and synthetic fuels. 9

Methodology 1st stage process Establish current and planned oil-based 1 2 and LNG bunkering infrastructures Identify supply of LNG export and import terminals 5 Isolate routes in stage 4 that trade in or out of ECA zones 4 Define high volume deep sea trade routes per ship type/size 3 Identify bunker ports in close proximity to ECA zones Derive voyage distances, average bunker consumption 6 and equivalent LNG 7 8 consumption per ship type/size Survey of shipowners to understand current options for mitigating emissions regulations Port survey to validate top 10 most likely LNG bunkering locations 1 LNG bunkering demand model 2nd stage process 2 Reality-based approach validation of findings by stakeholders Figure 3: The study steps 10

5. LNG bunker demand assessment shipowners survey We surveyed leading shipowners on deep sea trades to get their opinions on the options for compliance with the sulphur content limits. This enabled us to assess: 1. the likelihood of adoption of LNG as a fuel 2. the owners timelines for considering options for compliance with SOx emission regulations. The owners surveyed represented the largest within their respective shipping sector (as measured by cargo volumes shipped or ship numbers/tonnage). Survey findings Main bunkering ports for deep sea ships Figure 4 shows the main bunkering ports for deep sea ships of the owners surveyed. One third of the bunkering locations identified (9 of 28) are in confirmed ECAs. 20 15 10 5 0 Singapore Rotterdam Gibraltar No. of respondents Marseilles (Fos) Houston New York Hong Kong Cape Town & Durban Malta Fujairah Gothenburg (Skaw) Istanbul Seattle Genoa West Africa Suez Canal Ras Tanura Tokyo Bay Hamburg San Francisco & LA Port Klang Shanghai Pireaus Las Palmas Antwerp Busan Novorossiysk Vancouver Ports in confirmed ECAs Figure 4: Shipowners survey the main bunkering ports for deep sea ships Shipowners intentions for mitigation of SOx emissions Figure 5 shows the shipowners current intentions for mitigating SOx emissions. The findings, which consolidate survey findings for all ship types, can be summarised as follows: 1. Low-sulphur distillate fuel is seen as a short-term (within the next five years) solution. 2. Exhaust gas scrubbing is seen as a medium-term (five to 10 years) option to mitigate emissions. 3. Shipowners see LNG-fuelled engines as a viable option in the long term (10+ years), particularly for ships on liner trades (see figures 6 and 7). 11

30 25 LNG-fuelled engine Dual fuel Scrubbers Distillate Number of respondents 20 15 10 5 0 Short term Medium term Long term No intention Don t know Figure 5: Shipowners survey intentions for mitigating SOx emissions 12 LNG fuel Dual fuel Scrubbers MGO 10 Number of respondents 8 6 4 2 0 Short term Medium term Long term No intention Don t know Figure 6: Shipowners survey intentions of cruise ship owners for mitigating SOx emissions 6 LNG fuel Dual fuel Scrubbers MGO 5 Number of respondents 4 3 2 1 0 Short term Medium term Long term No intention Don t know Figure 7: Shipowners survey intentions of container ship owners for mitigating SOx emissions For other deep sea ship types there is doubt as to which compliance option would be best. For example, the majority of tanker owners who replied to the survey indicated that they don t know what mitigating technologies they would use to deal with the SOx emission regulations. 12

6. LNG bunker supply assessment port survey Twenty-five deep sea bunkering ports were surveyed about the likelihood of them developing LNG bunkering infrastructure. The survey was carried out to assess: 1. their awareness of LNG as fuel as an option for SOx emission regulation compliance 2. the plans of ports for developing LNG bunkering infrastructure. The ports surveyed were specially selected based on an assessment of the likelihood of becoming a future LNG bunkering port. Details of the survey, including summaries of questionnaire responses, are shown in Appendix 1. Survey findings Awareness of LNG as fuel as a compliance option Five out of 13 bunkering ports saw the use of LNG as a bunker fuel in deep sea shipping as very likely to happen. From the ports perspective the key drivers for a change to LNG are: 1. pricing and availability 2. local regulations a small number of responding ports have regulations in place for LNG bunkering (see Figure 8) 3. overcoming operational issues such as safety nine out of 13 ports said there will be restrictions in place for LNG bunkering. For a port to provide LNG bunkering facilities, the highest-rated commercial advantage is the port s location relative to an ECA and the implied demand that would bring. Yes 8% No 92% Figure 8: Port survey proportion of ports with local LNG bunkering regulation in place 13

Plans for LNG bunkering infrastructure The majority of responding ports do not have LNG bunkering development plans. Those ports seriously considering implementing LNG bunkering facilities see the funding coming from the private companies operating within the port. Where an LNG import terminal exists, or is being developed close by, most ports see the importing terminal as a key driver of providing LNG in small parcels for bunkering operations. European ports have carried out the most work and research into LNG as a fuel and the provision of LNG bunkering facilities. Consequently, they have a clearer view that LNG bunkering is likely to happen, starting with short sea shipping in ECAs and possibly cascading eventually into deep sea trade facilitated by regulations. However, the key driver demand is highly dependent on LNG pricing and its comparable price difference with competing fuels, for example HFO and MGO. The general consensus is that liner services would be particularly suited for LNG bunkering. This response aligns with the responses from the shipowners survey, where LNG-fuelled engine options were favoured in the medium/long term by cruise ship owners, followed by container ship owners. Table 3 illustrates the ports response to developing infrastructure and shows the year that either ECA or global sulphur limits will come into effect in each port. Port Developing LNG bunkering infrastructure? Year that ECA or global sulphur limit takes effect Gothenburg Yes Current (in port area only) Southampton No Current Zeebrugge Yes Current Algeciras No 2020 Rotterdam Yes Current Piraeus No 2020 Nynäshamn Yes Current Fujairah No 2020 Houston No Current Vancouver No Current Singapore Yes 2020 Shanghai No 2020 New York No Current Table 3: Analysis of the ports responses to providing LNG bunkering facilities 14

7. Forecast of LNG-fuelled newbuild and bunker demand We have made forecasts of LNG-fuelled newbuild and LNG bunker demand to 2025, considering LNG fuel pricing and regulation of fuel oil sulphur limits. LNG bunker demand model An interactive LNG bunker demand model has been developed as part of the study which enables the user to alter conditions based on a changing regulatory environment, technological developments or pricing of fuel and equipment. The outputs of the model are a view on demand for newbuilds adopting LNG dual-fuel operation up to 2025 for deep sea trade. The model also allows factors driving LNG bunkering demand and supply to be examined as a function of: regulatory pressure to burn cleaner fuels either within designated waters (ECAs) or globally availability of LNG bunkers at key trading and known fuel oil bunkering ports cost comparison of compliance options, average annual bunker costs and equipment costs known to date size of deployment of newbuilds on selected trade routes. The model comprises seven interactive workbooks in a spreadsheet created from a set of assumptions outlined in Appendix 2. The main workbooks are: ship types one workbook per ship type (container ships; cruise ships; dry bulk carriers; oil tankers) LNG bunkering costs LNG bunkering demand overview. The model includes a propensity figure for the adoption of LNG as a fuel for newbuilds based on the % voyage time expected in ECAs for the selected main deep sea trading routes (see Appendix 2). It is assumed that the greater the time spent in emissions-limited waters (whether in an ECA or globally after IMO regulations are implemented) the greater the propensity to choose LNG as a fuel for compliance with low emissions regulations. Additionally, a second layer of propensity to adopt LNG-fuelled engines is applied in the model based on comparable cost saving on LNG bunker prices compared with the cheapest alternative fuel options. Additional model outputs are contained in Appendix 2, with further breakdown and analyses by ship type. 15

The three demand scenarios: base case, high case and low case Three possible scenarios have been developed base case, high case and low case considering the input assumptions shown in Table 4. Factors Regulatory compliance Newbuild demand Fuel prices Input assumptions Emission control areas (ECAs) 0.5% global sulphur date Propensity to select LNG fuel from 2020 HFO/MDO/MGO 2012 forecast LNG bunker price forecast 2012-2025 Base case Confirmed ECAs High case - 25% Confirmed ECAs plus speculative ECAs in 2018 (Japan, Singapore, Panama) 2020 2020 2023 Increase propensity by 50% By year-on-year change in crude oil price By HFO/Henry Hub gas prices (75%/ 25%) yearon-year change Increase propensity by 75% Same as base case Low case +25% Confirmed ECAs Increase propensity by 25% Same as base case Base case - 25% Base case +25% Table 4: Input assumptions for the three scenarios base, high and low case LNG bunker and newbuild forecast: base case scenario LNG-fuelled newbuild forecast: base case scenario A base case scenario propensity for LNG-fuelled engines was set at a maximum of approximately 30% of global newbuilds prior to 2020 global sulphur limits, followed by an increase to a maximum of 45% post 2020 when a marked shift to proven technologies for emission compliance could be expected (of which LNG fuel is expected to be one). A total of 653 LNG-fuelled newbuilds has been forecasted using the model for the period up to 2025; this represents 4.2% of total global deliveries expected during the period. Newbuild deliveries and forecast demand for LNG fuelled newbuilds by ship type is illustrated in Table 5. In terms of ship type penetration, cruise ships showed the highest level of uptake of LNG as a fuel at 10.9% (25 ships) by 2025. Of the expected total global fleet of LNG-fuelled ships, 42% (275 ships) are expected to be dry bulk carriers, due to the size of global dry bulk carrier deliveries expected during the period. 16

Deep sea ship types Cumulative forecast newbuild deliveries (2012-2025) Cumulative forecast LNG-fuelled newbuilds (2012-2025) LNG-fuelled newbuilds as a % of global newbuilds Container ships 1,898 110 5.8% 16.8% Dry bulk carriers 7,305 275 3.8% 42.1% Oil tankers 1,977 146 7.4% 22.3% Cruise ships 230 25 10.9% 3.8% Chemical tankers* 1,614 14 1% 2.1% LPG tankers* 522 4 1% 0.7% General cargo ships* 1,313 49 3.8% 7.6% Car carriers* 711 30 4.2% 4.6% Grand total 15,570 653 4.2% 100% * Based on selected deep sea ship types with similar trading operation Table 5: Global newbuild forecasts versus LNG-fuelled newbuild demand base case (2012-2025) % market share per shiptype of all LNG fuelled newbuilds Global LNG bunker demand forecast: base case scenario For the base case scenario, global LNG bunker demand for deep sea trades looks comparatively small as a percentage of global HFO bunker demand, reaching approximately 24 MnT by 2025 (0.8% of global HFO bunker consumption by 2025) (see Table 6). Deep sea ship types Cumulative HFO bunker consumption MnT (2012-2025) Cumulative LNG bunker consumption MnT (2012-2025) Container ships 805 10.7 1.3% Dry bulk carriers 1,178 5.1 0.4% Oil tankers 352 2.6 0.7% Cruise ships 179 3.3 1.8% Chemical tankers* 75 0.0 0.0% LPG tankers* 20 0.0 0.1% General cargo ships* 83 0.2 0.3% Car carriers* 123 1.0 0.8% Grand total 2,815 24 0.8% * Based on selected deep sea ship types with similar trading operation Table 6: Marine fuel bunker demand forecasts LNG versus HFO base case (2012-2025) LNG Bunker as a % of global HFO bunker demand per ship type 17

600 Global LNG production Global LNG bunker consumption (MnT) per year LNG bunker as a % of global LNG production (RH axis) 1.8% Million tonne 500 400 300 200 100 1.6% 1.4% 1.2% 1.0% 0.8% 0.6% 0.4% 0.2% LNG bunker as a % of global LNG production 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0% Figure 9: Global LNG production versus LNG bunker demand 2012-2025 base case As we approach 2025, the volume of LNG bunker fuel is expected to reach 1.5% of global LNG produced per year for the base case scenario (see Figure 9). For the base case scenario, LNG bunker demand is expected to grow progressively to 2025 (see Figure 10), illustrating absolute LNG versus HFO/distillates demand per year. Global HFO bunker consumption (MnT) LNG bunker as a % of global HFO bunker demand (RH axis) Global LNG bunker consumption (MnT) 300 3.5% 250 3.0% Million tonne 200 150 100 50 2.5% 2.0% 1.5% 1.0% 0.5% LNG bunker as a % of HFO bunker 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0.0% Figure 10: Global bunker consumption all deep sea ship types LNG bunker versus HFO base case 18

LNG bunker price forecasts base case scenario Figure 11 illustrates our base case bunker fuel prices, driven by current market sentiment forecasts provided by MSI. 1,400 LNG-fuelled deliveries - base (RH axis) HFO Effective LNG (North America) LNG (Asia) LNG (Europe) 0.1% MGO 700 1,200 1,189.8 600 US $ per Tonne 1,000 800 600 895.7 500 400 300 number of ships 400 200 200 100 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0 Figure 11: Global forecasts fuel bunker prices interplay (2012-2025) base case ( HFO Effective is HFO with any variant of sulphur content higher than ECA or global limits at the given period.) Although LNG regional prices are still maintained at lower prices compared with MGO and HFO prices during 2020-2025, the demand seen is still relatively low. If shipowners can obtain LNG fuel cheaper than HFO or MGO and at a less volatile price with lower price differential regionally, they may be more convinced to shift to LNG as the preferred fuel in the future. Investigation of this interplay of price and take-up of LNG for bunker fuel is the basis of the following high case and low case scenario forecasts. LNG bunker and newbuild forecast: high case scenario LNG-fuelled newbuild forecast high case scenario The high case scenario is driven by the set of assumptions laid out in Table 4. The overall LNG-fuelled newbuild demand generated is expected to amount to 1,963 ships by 2025. The key driver for improvements in the LNG-fuelled newbuild demand is the 25% decrease applied to the current 2012 LNG bunker prices used in our base case model and the 75% increase applied to the propensity for newbuilds to convert to LNG-fuelled designs from 2020-2025. Sensitivity testing of the model indicates that the main reason for increased adoption of LNG fuel is the relative fuel cost saving between LNG regional bunker prices and the most cost-effective alternative fuel option. Consequently, it can be concluded that for LNG-fuelled newbuilds to be favoured, future LNG bunker prices will have to be kept at a much lower level compared to 0.1% MGO prices and HFO prices. Results for the high case scenario for LNG-fuelled newbuilds by shiptype per year are illustrated in Table 7. 19

Deep sea ship types Cumulative forecast newbuild deliveries (2012-2025) Cumulative forecast LNG-fuelled newbuild (2012-2025) LNG-fuelled newbuilds as a % of global newbuilds Container ships 1,898 277 14.6% 14.1% Dry bulk carriers 7,305 890 12.2% 45.4% Oil tankers 1,977 466 23.6% 23.8% Cruise ships 230 36 15.7% 1.8% Chemical tankers* 1,614 44 3% 2.2% LPG tankers* 522 14 3% 0.7% General cargo ships* 1,313 160 12.2% 8.2% Car carriers* 711 75 10.5% 3.8% Grand total 15,570 1,962 12.6% 100% * Based on selected deep sea ship types with similar trading operation Table 7: Global newbuild forecasts versus LNG-fuelled newbuild demand high case (2012-2025) % market share per ship type of all LNG fuelled newbuilds Similarly to our base case scenario, there is a strong adoption within the dry bulk carrier and tanker newbuild sectors; this may be attributed to the fact that, generally, high levels of global deliveries are expected within these segments compared with other ship segments. However, when the factor of uptake within each segment s global deliveries is considered, the oil tanker fleet demonstrates the highest uptake of 23.5%, followed by cruise ships with 15.7%, and container ships with 14.6%. Global LNG bunker demand forecast high case scenario The high case scenario is a significant improvement in LNG bunker demand, as illustrated in Table 8. Deep sea ship types Cumulative HFO bunker consumption MnT (2012-2025) Cumulative LNG bunker consumption MnT (2012-2025) Container ships 805 28.1 3.5% Dry bulk carriers 1,178 19.9 1.7% Oil tankers 352 8.9 2.5% Cruise ships 179 5.6 3.1% Chemical tankers* 75 0.1 0.2% LPG tankers* 20 0.1 0.3% General cargo ships* 83 0.9 1.1% Car carriers* 123 2.3 1.9% Grand total 2,815 65.8 2.3% * Based on selected deep sea ship types with similar trading operation Table 8: Marine fuel bunker demand forecasts LNG versus HFO (2012-2025) high case LNG Bunker as a % of global HFO bunker demand per ship type 20

Global LNG bunker demand could reach 65.8 MnT by 2025 in our high case scenario, representing a 173.1% increase from our base case scenario for LNG demand, 4.2% of global LNG production and 8% of global bunker demand at 2025 (see Figures 12 and 13). 600 Global LNG production (MnT) Global LNG bunker consumption (MnT) per year LNG bunker as a % of global LNG production (RH axis) 5% Million tonne 500 400 300 200 100 4% 3% 2% 1% LNG bunker as a % of global LNG production 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0% Figure 12: Global LNG production versus LNG bunker demand 2012-2025 high case 250 Global HFO bunker consumption (MnT) Global LNG bunker consumption (MnT) LNG bunker as a % of global HFO bunker demand (RH axis) 12% Million tonne 200 150 100 50 10% 8% 6% 4% 2% LNG bunker as a % of HFO bunker 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0% Figure 13: Global bunker consumption all deep sea ship types LNG bunker versus HFO high case LNG bunker price forecasts high case scenario Figure 14 illustrates the LNG bunker, HFO and MGO prices used for high case scenario development with forecast LNG-fuelled newbuild demand super-imposed. During the forecast period, LNG bunker prices regionally are extremely favourable, with HFO and MGO bunker prices expected to continue to increase and remain volatile over the period. 21

1,400 LNG-fuelled deliveries demand - high (RH axis) HFO Effective LNG (North America) LNG (Asia) LNG (Europe) 0.1% MGO 2,500 1,200 1,189.7 2,000 1,000 US $ per Tonne 800 600 671.7 1,500 1,000 number of ships 400 200 500 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0 Figure 14: Global forecasts fuel bunker prices interplay (2012-2025) high case LNG bunker and newbuild forecast: low case scenario LNG-fuelled newbuild forecast low case scenario The low case scenario is driven by our low case assumptions laid out in Table 4. Overall LNG-fuelled newbuild generated is forecast to only amount to 13 ships by 2025 in the low case scenario, as illustrated in Table 9. Deep sea ship types Cumulative forecast newbuild deliveries (2012-2025) Cumulative forecast LNG-fuelled newbuilds (2012-2025) LNG-fuelled newbuilds as a % of global newbuilds Container ships 1,898 5 0.3% 38.5% Dry bulk carriers 7,305 0 0.0% 0.0% Oil tankers 1,977 0 0.0% 0.0% Cruise ships 230 7 3.0% 53.8% Chemical tankers* 1,614 0 0% 0.0% LPG tankers* 522 0 0% 0.0% General cargo ships* 1,313 0 0.0% 0.0% Car carriers* 711 1 0.1% 7.7% Grand total 15,570 13 0.1% 100% * Based on selected deep sea ship types with similar trading operation Table 9: Global newbuild forecast versus LNG fuelled newbuild demand low case (2012-2025) % market share per ship type of all LNG fuelled newbuilds 22

There are two key factors limiting the size of demand for LNG-fuelled newbuilds: the 25% increase applied to LNG bunker prices used in our base case model the implementation year of the global sulphur limit shifting from 2020 to 2023 (further sensitivity testing indicated that shift of the year of implementation to 2025 generated a zero LNG-fuelled newbuild demand up to 2025). In the low case scenario forecast, demand builds up progressively from 2023 onwards. In terms of ship type uptake, the majority (58% seven ships) are expected to be cruise ships, which have a higher probability of encountering ECAs. Global LNG bunker demand forecast: high case scenario The low case scenario generated minimal LNG bunker demand when compared with the base case. Global LNG bunker demand is expected to reach 0.7 MnT by 2025, representing a 93% decrease from our base case scenario. This 0.7 MnT LNG demand also represents a mere 0.001% and 0.002% of global LNG production and HFO consumption demand at 2025 respectively (see Table 10 and Figures 15 and 16). Deep sea ship types Cumulative HFO bunker consumption MnT (2012-2025) Cumulative LNG bunker consumption MnT (2012-2025) LNG Bunker as a % of global HFO bunker demand per ship type Container ships 805 0.1 0.0% Dry bulk carriers 1,178 0.0 0.0% Oil tankers 352 0.0 0.0% Cruise ships 179 0.6 0.3% Chemical tankers* 75 0.0 0.0% LPG tankers* 20 0.0 0.0% General cargo ships* 83 0.0 0.0% Car carriers* 123 0.0 0.0% Grand total 2,815 0.7 0.0% * Based on selected deep sea ship types with similar trading operation Table 10: Marine fuel bunker demand forecasts LNG versus HFO (2012-2025) low case 23

600 Global LNG production (MnT) Global LNG bunker consumption (MnT) per year LNG bunker as a % of global LNG production (RH axis) 5% Million tonne 500 400 300 200 100 4% 3% 2% 1% LNG bunker as a % of global LNG production 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0% Figure 15: Global LNG production versus LNG bunker demand 2012-2025 low case 300 Global HFO bunker consumption (MnT) Global LNG bunker consumption (MnT) LNG bunker as a % of global HFO bunker demand (RH axis) 0.002 Million tonne 250 200 150 100 50 0.002 0.002 0.001 0.001 0.001 0.001 0.001 0.000 0.000 LNG bunker as a % of HFO bunker 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0.000 Figure 16: Global bunker consumption all deep sea ship types LNG bunker versus HFO low case 24

LNG bunker price forecasts low case scenario Low case scenario figures are quite pessimistic and driven by high LNG bunker prices over the forecast period. Assessment of LNG bunker prices indicates that if it is delivered at any price higher than USD 1,000 per tonne, or if LNG bunker prices are sold significantly higher than HFO and MGO prices, demand will be minimal. 1,400 LNG-fuelled deliveries - low (RH axis) HFO Effective LNG (North America) LNG (Asia) LNG (Europe) 0.1% MGO 13 14 1,200 1,189.7 12 1,000 1,119.5 10 US $ per Tonne 800 600 8 6 number of ships 400 4 200 2 0 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0 Figure 17: Global forecasts fuel bunker prices interplay (2012-2025) low case 25

8. Conclusions LNG as fuel is one option for deep sea shipping to comply with future emission regulations. From surveys of shipowners on deep sea trades and bunkering ports, as well as modelling of LNG fuelled newbuild and bunker demand, we have drawn the following conclusions: 1. LNG-fuelled engines are a viable option for deep sea trades in the long term (10+ years), particularly for ships on liner trades. This conclusion can be drawn from both the shipowner survey as well as the bunkering port survey. 2. Considering the base case scenario model, with what we know today about the factors affecting adoption of LNG, 653 newbuilds are expected to adopt LNG-fuelled engines by 2025 on deep sea routes. This represents 4.2% of global newbuilds forecast to be delivered during the period 2012-2025. 3. The high case scenario model output was much more favourable towards LNG-fuelled newbuilds when the forecast price of LNG bunker fuel was reduced by 25%. On the other hand, the low case scenario model with a higher forecast price of LNG bunker fuel and a later implementation date of global sulphur limits generated demand for just 13 LNG-fuelled newbuilds for deep sea shipping up to 2025. 4. LNG bunker demand is highly dependent on LNG pricing and its comparable price difference with competing fuels, e.g., current and future alternative fuels. This conclusion can be drawn from the bunkering port survey and is confirmed from outputs of the LNG demand model for LNG-fuelled newbuilds and LNG bunkering in the low case scenario. We will continue monitoring global commercial developments of LNG as a fuel, and provide annual updates of the forecasts of LNG fuelled newbuilds and LNG bunker demand. In addition, we will carry out ongoing validation and sensitivity studies for the model, updating it yearly and involving industry stakeholders including gas suppliers, engine makers, shipowners and shipyards. 26

9. References 1. IEA, World Energy Outlook Report, 2010. 2. Outlook for Marine Bunkers and Fuel Oil to 2030 A Key to Understanding the Future of Marine Bunkers and Fuel Oil Markets, published in July 2011 by Robin Meech, Marine and Energy Consultation Ltd. 3. Maritime Strategies International Ltd; Quarterly Shipbuilding Industry Forecast Standard ship Tables, February 2012; MSI Bunker forecasts obtained from MSI Tanker shipping proprietary model, February 2012, HFO bunker prices and consumption demand upto 2025. 10. Abbreviations and terms CO 2 carbon dioxide ECAs Emission Control Areas EGTS exhaust gas treatment system HFO heavy fuel oil HFO (effective) heavy fuel oil with any variant of sulphur content higher than ECA or global limits at the given period. IEA International Energy Agency LNG liquefied natural gas LSFO low-sulphur fuel oil MDO/MGO marine diesel oil/marine gas oil MSI Maritime Strategies International Ltd NOx nitrogen oxide PM particulate matter SOx sulphur dioxide 27

11. Figures and tables Figures Figure 1: Figure 2: Cumulative global LNG fuelled newbuilds and LNG bunker consumption (base, high and low cases). Page 5. MARPOL Annex VI regulations and enforcement of sulphur limits with respective timelines. Page 6. Figure 3: The study steps. Page 10. Figure 4: Shipowners survey the main bunkering ports for deep sea ships. Page 11. Figure 5: Shipowners survey intentions for mitigating SOx emission. Page 12. Figure 6: Shipowners survey intentions of cruise ship owners for mitigating SOx emissions. Page 12. Figure 7: Shipowners survey intentions of container ship owners for mitigating SOx emissions. Page 12. Figure 8: Port survey proportion of ports with local LNG bunkering regulation in place. Page 13. Figure 9: Global LNG production versus LNG bunker demand 2012-2025 base case. Page 18. Figure 10: Global bunker consumption all deep sea ship types LNG bunker versus HFO base case. Page 18. Figure 11: Global forecasts fuel bunker prices interplay (2012-2025) base case. Page 19. Figure 12: Global LNG production versus LNG bunker demand 2012-2025 high case. Page 21. Figure 13: Global bunker consumption all deep sea ship types LNG bunker versus HFO high case. Page 21. Figure 14: Global forecasts fuel bunker prices interplay (2012-2025) high case. Page 22. Figure 15: Global LNG production versus LNG bunker demand 2012-2025 low case. Page 24. Figure 16: Global bunker consumption all deep sea shiptypes LNG bunker versus HFO low case. Page 24. Figure 17: Global forecasts fuel bunker prices interplay (2012-2025) low case. Page 25. Figure 18: World map showing the ports surveyed. Page 30. Figure 19: Top trade routes by volume of global ship movements. Page 41. Figure 20: World fleet development by ship type 1991-2025. Page 46. Tables Table 1: Current and confirmed emission control areas. Page 7. Table 2: The three main options for compliance and corresponding emission reductions. Page 8. Table 3: Analysis of the ports responses to providing LNG bunkering facilities. Page 14. Table 4: Input assumptions for the three scenarios base, high and low case. Page 16. Table 5: Global newbuild forecasts versus LNG-fuelled newbuild demand base case (2012-2025). Page 17. Table 6: Marine fuel bunker demand forecasts LNG versus HFO base case (2012-2025). Page 17. Table 7: Global newbuild forecasts versus LNG-fuelled newbuild demand high case (2012-2025). Page 20. Table 8: Marine fuel bunker demand forecasts LNG versus HFO (2012-2025) high case. Page 20. Table 9: Global newbuild forecast versus LNG fuelled newbuild demand low case (2012-2025). Page 22. Table 10: Marine fuel bunker demand forecasts LNG versus HFO (2012-2025) low case. Page 23. Table 11: Top 10 global bunker locations by HFO/distillates bunker throughput at 2010. Page 29. Table 12: Tier 1 and Tier 2 bunker ports for survey. Page 30. Table 13: Main trade routes and areas per shiptypes selected for deep sea trades. Page 41. Table 14: Newbuild deliveries % expected LNG fuelled/dual fuel engine based on voyage time in Table 15: ECA and sulphur-limited water (base case, high case, low case scenarios). Page 42. Regional LNG import and export terminals, including number of oil bunkering terminals. Page 43. Table 16: LNG bunker cost savings and respective applied % demand for newbuilds. Page 44. Table 17: HFO, MDO & LNG regional bunker prices forecasts scenario 2012 applied prices. Page 45. 28

Appendix 1 Port survey analysis data and results Top 10 global bunkering locations Singapore, North West Europe and the Middle East account for about half of bunkering throughput globally. As shown in Table 11, Singapore is the largest of the global bunkering ports and is located on the main trade lanes between Europe, Middle East and the Far East. These top 10 ports account for almost 40% of the global bunkering volume throughput. Port Throughput (in thousand tonnes) Market share Singapore 39,011 17% Rotterdam 13,000 6% Fujairah 9,500 4% Antwerp 6,108 3% Hong Kong 5,429 2% Gibraltar 5,047 2% Korea (Busan) 4,559 2% West Africa 4,100 2% Tokyo Bay 3,494 2% Iran 3,135 1% Rest of world 138,530 60% Grand total 231,913 100% Table 11: Top 10 global bunker locations by HFO/distillates bunker throughput at 2010 Selection of bunkering ports for survey Twenty-five bunkering ports were selected for the survey based on specific criteria. According to which criteria they met, the ports were split into two groups tier 1 and tier 2. Six of the top 10 bunkering ports were included in tier 1. The ports are listed in Table 12 and Figure 18. Selection criteria Tier 1 ports that fulfilled up to two of the following criteria: 1. They are known bunkering ports. 2. They are known to be looking at their potential as an LNG bunkering site. 3. The supply of LNG is close to the port (within a 50-mile radius) or in the port area, which means there is availability of supply. 4. The port is located along a main deep sea trade route. Tier 2 ports that fulfilled other specific criteria: 1. Those considered as early adopters: Nynäshamn, Sweden. 2. Bunkering ports that ships may be able to deviate to before entering an ECA, for example Las Palmas in the Canary Islands. 3. Trade volume-specific bunkering ports for particular ship types, for example transit ports of dry bulk carriers and container ships through Gladstone and Sydney ports respectively. 29

Tier 1 ports Map reference Port Map reference Port A Singapore* B Rotterdam* C Fujairah* D Hong Kong E Algeciras* F Busan G San Francisco H Los Angeles I New York* J Panama Canal Cristobal K Panama Canal Balboa L Houston* M Gothenburg* N Piraeus* O Suez Canal Port Said P Shanghai* Q Zeebrugge* R Southampton* S Tokyo Bay Yokohama Tier 2 ports Map reference Port Map reference Port T Vancouver* U Las Palmas V Nynäshamn* W Kochi (Cochin) X Gladstone Y Sydney *Ports that responded to the survey Table 12: Tier 1 and Tier 2 bunker ports for survey Figure 18: World map showing the ports surveyed (Port K, Panama Canal Balboa, is behind port J.) 30

Additional notes on specific port selection Las Palmas is considered a deviation port as it is a top bunkering location for vessels on trade routes to and from Europe, and should be considered in terms of its deviation port potential for LNG bunkering. This could be a dedicated port/area along high volume trading routes where a ship can divert just for bunkering purposes. Nynäshamn is situated south of Stockholm, Sweden, and is an early adopter port; the port already has an LNG terminal and is looking at LNG bunkering facilities. The potential to provide LNG bunkering facilities in the region has been thoroughly investigated by the port. As a result, the terminal is ready to commence scalable LNG bunkering services as demand increases. Kochi (Cochin) is a port that is looking to become a bunkering hub and also to compete with other ports along the Asia to Europe trade lane. They are developing the Multi User Liquid Terminal (MULT) at Puthuvypeen SEZ (International Bunkering Terminal). Gladstone in Australia is a major port for the shipment of coal and the port also provides bunkering facilities. The Australian government will be implementing a carbon tax with the aim of reducing the country s carbon emissions. This will affect the operations of major commodities companies which may lead to companies considering ways to reduce their carbon emissions through the supply chain, with shipping being a component of this. Therefore, if the main users of the port are looking at LNGfuelled ships, the port may consider providing the LNG bunkering facilities. Sydney is included as an Australian port with significant volume of different cargo trades through the port. Some of the ports above were included in the full survey due to the fact that there were strong indications and information provided publicly that they were looking at LNG bunkering. An example is Vancouver. Port profiles Each port s profile provided important background information for assessing the survey responses. Nine ports of those that responded are port authorities that act as landlord ports. In terms of investment, the ports provide the land (including road networks, quay walls, jetties, etc.) and the operators invest in facilities (infrastructure development pipelines, vessels, storage facilities, etc). Nevertheless, it is the port authority that would develop and enforce the regulations for LNG bunkering and make the decision as to where and how LNG bunkering operations would be carried out within the port area. One of the European landlord ports said they are involved in every stage of the planning and development of LNG bunkering within the port. For one of the European ports, the survey was completed by the LNG terminal operator within the port. The terminal operator is already an LNG export terminal and will be providing LNG bunkering to a shipping line (once the vessels are built and delivered). At two European ports, the operators/shipowners (users) that will start using LNG-as-a-fuel bunkering facilities are incentivised by the government. Some of the port authorities, particularly those that have not carried out research into LNG, are quite uncertain about LNG as a fuel and this is reflected in the way they responded to the survey questions. The specific people who completed the survey included heads of planning and environment departments, and heads of marine services divisions. In some cases the survey was passed around several departments within the ports to ensure the questions were answered as accurately as possible. 31

The survey Part 1 Awareness: To gauge ports awareness of LNG as a bunker fuel and how they see themselves relative to its development. Question 1: Are you aware of the potential implications of the IMO MARPOL Annex VI regulations on fuel type availability and price? Yes No Number of ports 13 0 Question 2: Generally, do you consider LNG as a bunker fuel is viable for deep sea shipping within the next 10-15 years? Very likely Somewhat likely Neutral Somewhat unlikely Very unlikely Number of ports 13 0 Port comments: one Asian port responded somewhat unlikely and commented: on the basis that dual-fuel technology for internal combustion engine limits foregoing steps. Question 3: What do you think will drive change to the use of LNG as a fuel from a port s perspective? Rating Pricing Supply availability Overcoming operational issues Increased reliability of ship from LNG Establishment of local regulations Positive public perception Other factors 1 High 11 8 4 1 3 1 N/A 2 Medium 1 2 7 5 4 2 N/A 3 Low 1 3 3 5 4 6 N/A 4 Very low 0 0 0 2 2 4 N/A Question 3a: Do you see yourself (the port) as a driver of change regarding the use of LNG as a fuel? * Based on 13 responses No 38% Yes 62% 32

Question 3b: Does your port already have local regulations governing the bunkering and/or operation of LNG-fuelled ships? Yes 8% * Based on 13 responses No 92% Question 3c: Are there any, or are there likely to be any, restrictions in place in your port as to when and/or where a ship may bunker LNG as fuel, in addition to those generally applicable to bunkering? Yes No Number of ports 9 4 Port comments: Restrictions/recommendations on bunker vessel design; Restrictions on bunker locations/operations. A certified operational procedure would be needed in order to guarantee safety. Ensuring safe distance around LNG operation. Obtaining construction/operating permits due to federal and state (concerns) regulations. Bunkering protocols primarily aim to prevent environmental pollution due to accidental oil spills. While this may not be a high priority issue when transferring LNG, certain safety protocols will likely need to be developed. These may be developed as industry standards but certain requirements may be port specific. The object of such regulation may be the technology used and the location (anchorage inside/outside, alongside, vicinity to residential or other sensitive areas etc). Lack of LNG fuel station or ship. Question 4: How would the port rate the importance of the potential commercial advantages of providing bunkering facilities for an alternative fuel type such as LNG? 14 High Medium Low Very low 12 Number of ports 10 8 6 4 2 0 A: Regulation compliance B: Positive public perception C: Location near an ECA with implied demand D: Differentiator to other ports 33

Question 5: In July 2011, the International Association of Ports & Harbours (IAPH) launched a project to develop guidelines for LNG bunkering in ports. Is your port participating in this project? Yes No Number of ports 1 12 Question 6: Has your port commissioned either internal or external research into LNG bunkering infrastructure development, planning, etc? * Based on 13 responses No 46% Yes 54% Observation: Most of the ports that responded No were identified to have independent bunker suppliers within the ports, and subsequently expected to make the decisions on LNG bunkering not the port authority. Part 2 Planning for LNG bunkering: To determine which ports have taken further steps into planning for LNG bunkering. Question 1: What type of bunkering facilities are currently provided in your port for deep sea shipping? (Including ferries, if applicable) Barge Road tanker Pipelines at berths Other Comments Yes 10 5 4 1 Bunker ships No 1 5 6 9 N/A 1 1 1 1 Question 2: In your port Master Plan, what priority is LNG bunkering given relative to other port development projects and why that level? High priority Medium prirority Low priority Not in Master Plan Why? Number of ports 3 1 0 8 See port comments 34

Port comments: Issue is too new. LNG bunkering will be an issue in the future. Current Master Plan was issued in 2000 and revised in 2004. It considers new infrastructures for bunkers (nowadays under construction), not specifically for LNG, but easily adaptable. The whole planning system is being updated, beginning with the Strategic Plan where, at least, medium priority will be specifically given. A real demand is being considered for small scale LNG (ferries). LNG bunkering project is considered as a commercial project, so it is much more short-term than is required in a Master Plan; Business plans reviewed annually based on analysis and customer/industry input; at this time, LNG bunkering has not been identified as an emerging issue in the long-term strategy from a business or port development perspective. 25% No 67% 8% High priority Medium priority Not in Master Plan Question 3: If the provision of LNG bunkering is a high or medium priority (Question 2) what timeframe for LNG bunkering operations to commence are you expecting? 0-5 yrs 5-10 yrs 10-20 yrs Beyond 20 yrs Number of ports 3 3 1 0 No answer (N/A) 6 6 6 6 Question 4: If the provision of LNG bunkering is a low priority or not included in the Master Plan (Question 2) what factors would improve its priority in terms of its importance to the port? Financial support Government policy Availability of supplier to supply port Other factors Comments 1 Most important 0 2 2 3 2 High importance 5 3 2 0 3 Medium 0 1 0 1 4 Low importance 0 0 1 0 See port comments 5 Least important 1 0 1 0 No answer (N/A) 5 5 5 6 35

Port comments: Other factors mentioned were: demand private investors at LNG demand side private investors at operational site (bunker company) increase of world fleet using LNG as a bunker fuel interest of LNG fuel producers. Question 5: If an LNG bunkering infrastructure facility is being planned at your port, how does the port intend to fund the infrastructure development and what organisational structure do you envisage? Port comments: There is co-operation between the port and a power company owned by the City. The port have quay and land and the power company build the terminal. No answer: too early for comment. Existing facilities are easily adaptable for deep sea LNG bunker; therefore new facilities would be needed in the short/medium-term. Any new investment would be related to storage capacity and pipelines which are always private according to a landlord model. The port does not invest in the bunkering facilities. That is done by private parties. The port does invest in the general infrastructure for which the port builds business case. The port would want certain return on investment (ROI) for normal commercial projects. Volume or income guarantees are business as usual. LNG bunkering is considered to be of strategic importance. The LNG terminal plans to fund LNG bunkering privately. Technically LNG can be provided at the port from a local LNG import terminal. However, the economics seem more challenging where for example; a sufficiently robust business case will be needed to incentivise the port and a private bunker fuel supplier to invest in meeting long-term LNG demand. In the short-term, incentive programs launched by the port could promote new green technologies and pave the way for LNG in the future. Question 6: From where do you anticipate sourcing the LNG required for bunkering and in what volumes? Source Existing local suppliers, or from import LNG terminals in Europe Volume Up to 600,000 tonnes per year for bunkering and peak shaving (port in Europe) Existing LNG terminal within the port From several importing LNG terminals already available and from a new project being currently developed Not known yet Existing LNG terminal within the port Local supplier and other European terminals Less than 200,000 tonnes per annum International market Local supplier 36

Question 7: Are there any LNG import terminals operating or planned near your port? If Yes, please answer 7a. If No, move to 7b. Yes No Number of ports 6 6 Question 7a: Do you consider the importing LNG terminal (main supplier) is likely to facilitate and/or participate in the development of the supply infrastructure for the small parcels of LNG required for bunkering? Yes No N/A Number of ports 4 4 4 Question 7b: Which other industry stakeholders do you consider may facilitate development of LNG bunkering supply infrastructure at your port? Please provide details. Port comments: The local power company will use LNG for peak shaving for their power plant nearby. The port also knows that there is an interest from large global LNG producers to be part of the development of infrastructure. Municipality (strong support already), shipowners (in order to use public bids), the passenger and ferry service consumers/ individual companies etc, new generation of consumers. The countries Association of petroleum producers may take on an important role in marketing LNG as a marine transportation fuel. The current bunker fuel suppliers are the most likely candidates and also shipping lines. Question 8: What proportion of the total bunkering volume at your port do you expect to be in LNG within the following periods? 5 yrs 10yrs 15 yrs North European port 2-5% 30% 40% North European port 0-10% short-sea 0-20% short sea & deep sea Unknown North European port 10% (of inland shipping) 20% (of inland shipping) 40% North European port 10% 20% 50% North American port <1% <1% <1% Port comments: 6 ports responded Unable to comment (too early to say). 37

Question 9: What are the most important drivers for your port to provide LNG bunkers in terms of the port s commercial position? 12 Most important High importance Medium Low importance Least importance 10 Number of ports 8 6 4 2 0 Other competing bunkering ports Pricing of LNG vs other fuels Location relative to ECA Number of ship calls Provision for LNG bunkering facilities Demand from ship owners and suppliers Positive public perception Status as major bunkering port * Based on 11 responses. One respondent did not answer this question Question 10: How are existing bunkering operations currently carried out for different ship types at your port? Ship types By barge at berth By fixed pipeline at berth Ship is relocated to bunkering berth or anchorage Bunkering is carried out at General anchorage Mix of options Container ships 9 0 0 3 N/A Oil tankers 6 2 1 5 N/A Dry bulk carriers 9 1 1 4 N/A Passenger cruise 5 0 0 1 N/A Passenger ro-ro 6 0 0 2 N/A Question 11: How do you expect LNG bunkers to be supplied at your port for the following ship types: container ships, tankers, dry bulk carriers (above 10,000 dwt) and passenger cruise ships and passenger ro-ro ships (above 5,000 gross tonnes (gt)). 16 Number of ports 14 12 10 8 6 4 Mix of/other options Bunkering at general anchorage Ship relocated to bunkering berth or anchorage By fixed pipeline at berth Dry barge at berth 2 0 Container ships Oil tankers Dry bulk carriers Passenger cruise Passenger Ro-Ro ships 38

Port comments: Some ports will use up to three different types of bunkering operations per ship type and other options include fixed LNG line and tanker truck. Some ports said it is too early to say. Question 11c: Which ship types do you consider are best suited for LNG bunkering at your port and why? Port responses: Ro-Ro & ferries because of their frequent calls at the port. The port has not undertaken any research into LNG bunkering opportunities and therefore can provide only a limited response. Initial concern would be the health and safety regarding LNG. Deep sea vessels at the port are mainly container ships, pure car carriers, general cargo vessels and liquid bulk vessels. Ferryboats, as there is a current demand, and any other type of ship suitable to be provided by barge. Container and tanker, inland barges. Tanker due to percentage call and bunkering volume. An open question at the present time. Liner ships particularly those engaged in the transpacific trade due to better economies of using less expensive fuels and presumably the availability of LNG bunker ports in the transpacific trade area. Comparison of questions 10 and 11: Existing bunkering operations versus proposed LNG bunkering operation options. 10 Existing bunkering operation Proposed LNG bunkering operations 9 8 Number of ports 7 6 5 4 3 2 1 0 By barge at berth By fixed pipeline at berth Ship relocated to bunkering berth or anchorage Bunkering at general anchorage Mix of/other options Mix of/other options = fixed LNG line and tanker truck. Note: Existing bunkering operation (Q10) had 10 responses and Proposed LNG bunkering operations (Q11) had eight responses. 39

Appendix 2 LNG bunkering newbuild demand model Shipping trade route inputs The model is based on deep sea trading ships only. Specifically, it is based on four major ship types, with major representative trade routes selected for each one to examine in depth. See representative selected trade routes per ship type listed in Table 2. Subsequently, the findings are later used to extrapolate the full results on the entire ship type sector. The ship type categories and corresponding ship size ranges applied are: container ships (> 2,900 teu) cruise ships (> 5,000 gt) dry bulk carriers (> 25,000 dwt) oil tankers (> 35,000 dwt). Additionally, in order to obtain a much more robust trend of adoption on a global scale and derive a sense of uptake within other ship types on deep sea trades, appropriate ship types, similar in operational behaviour and trading patterns to the four major ship types above, have been used to derive the sector-specific results. The selected ship types are: chemical tankers (> 5,000 dwt; based on oil tanker fleet percent uptake) car carriers (> 5,000 dwt; based on container ship uptake due to their similarities in liner trading operations) LPG tankers (> 5,000 dwt; based on oil tanker fleet uptake, due to movements of liquid cargo parcels) general cargo ships (> 5,000 dwt; based on dry bulk carrier uptake; similar to container ship liner operations). The global deep sea trades for container ships, oil tankers, dry bulk carriers and cruise ships were assessed based on a simple tonne-mile calculator extracted from MSI s shipping economic models covering the container ship, dry bulk carrier and tanker sectors. For cruise ships, the movements of every cruise ship in the fleet were analysed to identify the top routes/operating areas (see Table 13). In order to make the development of the model practical, the main routes per ship type and size were selected with reference to the current ECA transit/proximity and likelihood of LNG bunker availability. Clearly, these are not the only accountable deep sea routes. For example, within the container ship sector there is likely to be significant new construction for deployment on the north-south trades and the feeder trades may also provide opportunities for deployment of LNG propulsion technology. Figure 19 Illustrates the volume of global ship movements in a one-year period showing the network of links between ports and regions. 40

Dry bulk carriers (> 25,000 dwt) Oil tankers (> 35,000 dwt) Container ships (> 2,900 teu) Cruise ships (> 5,000 gt) Australia to China Middle East to South Asia Far East to Europe Asia Europe to North America Western Europe to North America Transatlantic North Europe/Baltic Australia to Japan Middle East to China Transpacific Caribbean/Central America Europe to Asia Latin America to Europe West Africa to North America Middle East to Southern Europe Latin America Middle East North America (East & West Coast) Oceania Southern Europe/North Africa/ Eastern Mediterranean Table 13: Main trade routes and areas per shiptypes selected for deep sea trades Figure 19: Top trade routes by volume of global ship movements 41

Regulatory drivers The model details emission control areas (ECAs) by way of sulphur content limits and year of implementation per ECA with some speculative ECAs included for future use. The introduction of the IMO s global sulphur limits regulations is also factored in along with the flexibility to alter the assumptions within the model in case the implementation date shifts in the future. It is assumed that the greater the time spent in emission-limited waters (whether an ECA or globally after IMO regulations are implemented) the greater the propensity to choose LNG as a solution to meet low emission regulations. The user input assumptions are: 1. the global sulphur limit implementation year 2. percentage of voyage time spent in an ECA 3. percentage of time spent in sulphur-limited waters per year. Future newbuild propensities to adoption of LNG-fuelled/dual-fuel engines are calculated based on sets of rules as a function of user input assumptions 2 and 3 above (see Table 14). Scenarios 2012-2019 Base case High case Low case Voyage time spent in ECA/ Global sulphur limits Propensity of newbuilds (2012-2019) Propensity post 2020 Propensity post 2020 Propensity post 2023 <5% 0% 0% 0% 0% 5-24.9% 10% 15% 19% 11% 25-54.9% 20% 30% 38% 23% 55-74.9% 25% 38% 47% 28% 75-100% 30% 45% 56% 34% Cruise ships existing ECA Cruise ships global sulphur limit Factor of increase (Based on LR sentiment April 2012) 15% 23% 28% 17% 30% 45% 56% 34% 50% 75% 25% Table 14: Newbuild deliveries % expected LNG fuelled/dual-fuel engine based on voyage time in ECA and sulphur-limited water (base case, high case, low case scenarios) 42

LNG supply route assumptions 70% of LNG terminals identified globally have established oil bunkering facilities available within the port or nearby. This is highlighted within Table 15. The table shows the number of LNG import and export terminals regionally, including LNG terminals where oil bunkering facilities are currently available. Region/LNG terminals Import terminals Export terminals Oil bunkering facilities Grand total North America East (i.e., Canaport LNG; Texas; Georgia Elba Island) North America West (Kenai Port Alaska) North East Asia (Shanghai, Fujian; Shenzhen terminals) South East Asia (Arun LNG; MLNG Pita; Kochi terminals) North West Europe (Rotterdam Gate; Zeebrugge Fluxys) Latin America (Pecem LNG FSRU; Andres LNG; Atlantic LNG) 10 0 6 10 0 1 1 30 0 16 30 4 6 7 10 8 2 8 10 7 2 9 9 Oceania (Australia; Darwin LNG; Karratha North West LNG) 2 1 2 OECD Mediterranean (Huelva; Cartagena LNG; Izmir Aliaga LNG) 6 6 6 North Africa (ELNG IDCO; Sonatrach Plant) 5 5 5 Middle East (Rasgas I, II, & III) 2 7 7 9 West Africa (NLNG Bonny) 2 1 2 Grand total 67 27 66 94 Table 15: Regional LNG import and export terminals, including number of oil bunkering terminals 43

The second key factor is the availability of LNG bunker supply at key ports. For this, the model draws on information gathered within the port survey to develop a view of LNG supply at ports along the trade routes under review. This data includes the number of bunker ports in ECAs and the number of bunkering ports/facilities that will possibly provide LNG bunkering upto 2025. The LNG amount required per tonne of HFO consumption is also taken into account. For example, a conversion of 0.80 tonne of LNG = 1 tonne of HFO is used for all assumed daily fuel consumption per ship. It is also assumed that deep sea ships, even on conversion to an LNG-as-fuel concept, will operate on a dual-fuel engine arrangement. In such cases, about 5% of fuel oil will be used as fuel injection and has therefore been factored into the cost calculations for the fuel consumption required on an LNGfuelled engine. The user input assumptions are: number of LNG bunker ports approximate LNG volume MnT available at each ECA region/port/location at least a minimum of one bunker facility will be made available on deep sea routes by 2020. Vessel deployment and operational factors routes/vessels assumptions The future deployment of LNG-fuelled/capable vessels on each route is based on the number of newbuilds deployed/available to be deployed on that route. Average vessel speeds and engine power per vessel size and type are used to forecast bunker consumption per ship up to 2025. The user input assumptions are: number of vessel deployed on route. number of newbuildings deployed on route based on MSI standard ships forecasting model 2012Q1. vessel operating speeds, engine power per vessel type and size. Comparable fuel prices of LNG versus HFO with scrubber and MGO This is a check on the price of LNG that makes the whole concept of LNG as a fuel either viable or uneconomical; it attempts to provide a commercial rationale within the demand calculations. Forecasts of fuel prices up to 2025 were applied to derive the cost of fuel options per year. Saving by using LNG bunker compared with HFO/MGO <-10% 0% -9.9% to 0% 20% 0% to 9.9% 40% 10% to 19.9% 60% 20% to 29.9% 80% >30% 100% Propensity for newbuilds to convert to LNG-fuelled/dual-fuelled Table 16: LNG bunker cost savings and respective applied % demand for newbuilds 44

The user input assumptions are: Gas reference prices are Henry Hub natural gas prices provided by MSI 2012Q1 (LNG sector model) HFO/MGO prices driven by crude oil prices forecast provided by MSI 2012Q1 (Tanker sector model). Future HFO & MGO prices are kept at the same forecasted prices for all the scenarios. Regional LNG bunker fuel price forecasts upto 2025 Asia, North America, European, derived with a combined function of 25% of natural gas price annual change and 75% of HFO price annual change upto 2025. The rationale is that future LNG bunker prices will be highly influenced by other fuel option prices increasingly as we approach 2020 global sulphur limits. Fuel prices 2012 (average forecast prices) USD per tonne Base case High case (-25%) Low case (+25%) HFO (effective)* 702 702 702 MGO 0.1% 1,000 1,000 1,000 LNG bunker North America 474 356 592 LNG bunker Asia 948 711 1,185 LNG bunker European 664 498 829 Table 17: HFO, MDO & LNG regional bunker prices forecasts scenario 2012 applied prices (LNG regional prices base case: the prices are our best estimates (as of April 2012) based on various sources for LNG as bunker delivered on board (i.e. including handling and infrastructure costs)) *HFO (effective) is HFO with any variant of sulphur content higher than ECA or global limits at the given period. Engine and abatement technology cost assumptions The engine and equipment costs per compliance option are also taken into account to provide some level of CAPEX for each ship type and size range. The user input assumptions are: cost of standard engines per power unit (USD/kW) exhaust gas treatment system (scrubber) cost per power unit (USD/kW) average reference engine power (MW) per ship type size ranges derived dual-fuel main engine cost per power unit (USD/kW) LNG storage tank costs per ship type and size range fuel gas supply system and additional equipment costs per ship type and size range finance interest rate for the cost is applied at 5% per year World fleet development by shiptypes 1991-2025 Figures obtained from the IHS-Fairplay database at the end of March 2012 showed the world fleet of existing ships (including commercial, self-propelled ships above 99 gt) amounted to over 103,560 ships at 1,046 mgt. The current orderbook of ships within the same category stood at above 6,765 ships (209.7 mgt). 45

According to MSI s shipbuilding forecast report published in March 2012 (based on ships above 5,000 gt) considering the current market economic conditions, the world fleet is expected to grow by approximately 35% from 987 mgt at the end of 2011 up to 1,282 mgt by 2020 and a further 18% to 1,511 mgt by 2025, when the global sulphur limit commences (see figure 20). Cruise ships Ropax Ro-ros PCCs Reefers General cargo LNG ships LPG ships Container ships Tankers Bulk carriers % annual change (RH axis) 1,600 10 1,400 9 Million GT 1,200 1,000 800 600 8 7 6 5 4 % annual y-o-y change 400 3 2 200 1 0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 0 Figure 20: World fleet development by ship type 1991-2025 46

Global newbuild forecasts for deep sea trades 2012-2025 Cumulative newbuild deliveries per ship type 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Container ships 204 358 500 582 681 795 916 1,050 1,196 1,356 1,512 1,645 1,768 1,898 Dry bulk carriers 898 1,560 2,014 2,316 2,638 3,006 3,399 3,814 4,266 4,799 5,451 6,147 6,768 7,305 Oil tankers 191 373 463 515 599 707 855 1,029 1,237 1,433 1,601 1,735 1,859 1,977 Cruise ships 8 14 20 32 49 67 87 106 126 145 166 186 206 230 Chemical tankers 102 170 226 314 435 576 699 808 913 1042 1,195 1,356 1,493 1,614 LPG tankers 23 50 76 117 164 204 233 264 315 379 436 472 498 522 General cargo ships 159 275 351 413 474 542 619 710 821 942 1,056 1,149 1,229 1,313 Car carriers 30 56 87 129 164 192 232 306 395 490 571 643 687 711 Grand Total 1,615 2,855 3,737 4,417 5,205 6,088 7,040 8,088 9,270 10,587 11,988 13,333 14,508 15,570 47

Global newbuild forecasts for deep sea trades 2012-2025 Base case Global LNG-fuelled newbuilds 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Container ships 0 1 1 1 1 1 21 43 60 73 91 110 Dry bulk carriers 1 34 65 88 129 211 275 Oil tankers 0 0 14 30 61 91 118 146 Cruise ships 0 0 1 3 6 9 13 18 25 Chemical tankers 0 0 0 0 0 0 0 1 4 9 14 LPG tankers 0 0 0 0 0 0 0 0 1 3 4 General cargo ships 0 0 0 0 0 7 13 17 24 38 49 Car carriers 0 0 0 0 0 0 0 5 12 17 21 26 30 Grand Total 0 1 1 1 1 2 3 84 168 252 356 514 653 Cumulative newbuild deliveries 1,615 2,855 3,737 4,417 5,205 6,088 7,040 8,088 9,270 10,587 11,988 13,333 14,508 15,570 LNG fuelled deliveries as % of global deliveries 0% 0% 0% 0% 0% 0% 0% 0% 1% 2% 2% 3% 4% 4% 48

Global newbuild forecasts for deep sea trades 2012-2025 High case Global LNG-fuelled newbuilds 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Container ships 1 2 3 3 4 6 11 61 114 156 198 236 277 Dry bulk carriers 3 5 7 8 13 21 29 154 307 421 530 731 890 Oil tankers 1 1 1 1 1 1 1 70 151 226 296 381 466 Cruise ships 0 0 1 1 2 7 12 16 21 27 36 Chemical tankers 0 0 0 0 0 0 0 0 0 0 4 19 32 44 LPG tankers 0 0 0 0 0 0 0 0 0 0 2 7 11 14 General cargo ships 0 1 1 1 1 2 4 5 30 60 82 99 133 160 Car carriers 0 0 0 1 1 1 1 3 16 31 44 57 67 75 Grand total 7 10 13 15 21 35 52 337 676 951 1,227 1,618 1,963 Cumulative newbuild deliveries 1,615 2,855 3,737 4,417 5,205 6,088 7,040 8,088 9,270 10,587 11,988 13,333 14,508 15,570 LNG fuelled deliveries as % of global deliveries 0% 0% 0% 0% 0% 0% 0% 1% 4% 6% 8% 9% 11% 13% 49

Global newbuild forecasts for deep sea trades 2012-2025 Base case Global LNG-fuelled newbuilds 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Container ships 4.4 Dry bulk carriers Oil tankers Cruise ships 0 0 0 0 2 4 7.3 Chemical tankers 0 0 0 0 0 0 0 0 0 0 0 0 0 0 LPG tankers 0 0 0 0 0 0 0 0 0 0 0 0 0 0 General cargo ships 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Car carriers 0 0 0 0 0 0 0 0 0 0 0 0 0 1.2 Grand total 0 0 0 0 2 4 13 Cumulative newbuild deliveries 1,615 2,855 3,737 4,417 5,205 6,088 7,040 8,088 9,270 10,587 11,988 13,333 14,508 15,570 LNG fuelled deliveries as % of global deliveries 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.001% 0.001% 0.001% 0.001% 0.01% 0.03% 0.08% 50

Global LNG bunker consumption demand 2012-2025 Base case Global all deep sea ship types 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Total Demand (2012-25) Global HFO bunker consumption (MnT) 166.3 175.4 178.1 180.0 183.9 188.2 194.0 199.7 206.1 213.9 222.3 229.7 235.8 241.4 2,814.8 Global LNG bunker consumption (MnT) 0.0 0.0 0.0 0.0 0.0 0.0 0.1 1.1 2.2 3.2 4.2 5.6 7.7 24.1 LNG bunker as a % of global HFO bunker demand 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.5% 1% 1.4% 2% 2.4% 3.2% 0.9% % market share per ship type global LNG bunker demand 0.0% 0.0% 0.0% 0.0% 0.0% 0.1% 0.1% 0.2% 4.5% 9.1% 13.2% 17.5% 23.2% 31.9% 100% High case Global all deep sea ship types 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Total Demand (2012-25) Global HFO bunker consumption (MnT) 166.3 175.4 178.1 180.0 183.9 188.2 194.0 199.7 206.1 213.9 222.3 229.7 235.8 241.4 2814.8 Global LNG bunker consumption (MnT) 0.1 0.1 0.1 0.1 0.2 0.3 0.5 3.6 7.0 9.7 12.2 15.4 20.2 69.6 LNG bunker as a % of global HFO bunker demand 0.0% 0.0% 0.0% 0.1% 0.1% 0.1% 0.2% 0.3% 1.7% 3% 4.4% 5% 6.5% 8.4% 2.5% % market share per ship type global LNG bunker demand 0.0% 0.1% 0.1% 0.2% 0.2% 0.3% 0.5% 0.8% 5.2% 10.0% 13.9% 17.6% 22.1% 29.1% 100% 51

Global LNG bunker consumption demand 2012-2025 Low case Global all deep sea ship types 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 Total Demand (2012-25) Global HFO bunker consumption (MnT) 166.3 175.4 178.1 180.0 183.9 188.2 194.0 199.7 206.1 213.9 222.3 229.7 235.8 241.4 2,814.8 Global LNG bunker consumption (MnT) 0.0 0.0 0.0 0.0 0.1 0.2 0.4 0.7 LNG bunker as a % of global HFO bunker demand 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0% 0.0% 0% 0.1% 0.2% 0.0% % market share per shiptype global LNG bunker demand 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.9% 0.9% 0.9% 0.9% 13.1% 24.2% 59.3% 100% 52

For further information or to discuss topics related to LNG as a fuel for deep sea shipping, please contact one of the following Lloyd s Register or MSI specialists: Lloyd s Register Jesper Aagesen, Senior Surveyor, Ship Design Specialist jesper.aagesen@lr.org Latifat Ajala, Senior Marine Market Analyst latifat.ajala@lr.org Maritime Strategies International, Ltd Stuart Nicoll, Consultant stuart.nicoll@msiltd.com www.lr.org/lngasfuel August 2012 Lloyd s Register is a trading name of Lloyd s Register Group Limited and its subsidiaries. For further details please see www.lr.org/entities.